OBJECTIVE: to determine whether a low fat, high fiber, high vegetable and fruit eating plan will decrease the recurrence of adenomatous polyps of the large bowel. BACKGROUND: In the PPT, three dietary factors are being investigated simultaneously through intervention with a realistic, comprehensive dietary pattern change. Large bowel adenomas present a unique opportunity to conduct an intervention trial because of the high prevalence of these lesions, the high recurrence rate, and the fact that adenomas are considered precursors for most large bowel cancers. METHODS: Men and women with one or more adenomas recently removed were randomized at eight Clinical Centers between June 1991 and January 1994. The goals for the intervention group are 20% of calories from fat, 18g fiber/1000 kcal, and 5-8 servings of fruits and vegetables (the exact number based on caloric intake). The intervention program integrates nutrition education and behavior modification techniques. Participants in the control group are expected to maintain their usual diet. Three different dietary assessment instruments are used: a modified Block/NCI Food Frequency Questionnaire, Four Day Food Records, and 24 Hour Recalls. Participants undergo colonoscopy again at one (T1) and four (T4) years into the study. Histologic type and extent of atypia are determined by central pathologists for all resected lesions. Blood specimens, collected annually, are analyzed for lipids, carotenoids, and vitamins A and E, and are available for hormonal and molecular genetic analyses. Rectal mucosal biopsy specimens are obtained at three time points for approximately 300 participants. PROGRESS: 2,079 were randomized into the study, with 1,037 in the intervention and 1,042 in the control group. The mean age of PPT participants at baseline was 61.5 years, with 35% women. At study entry the mean % calories from fat was 36%, mean fiber intake was 9.5 g/1000 kcal, and mean number of fruit and vegetable servings/day was 3.8. At baseline 35% of participants had two or more adenomas, 29% had one or more large (greater than or equal to 1 cm) adenomas. Demographic, nutritional, and pathologic characteristics were comparable within the intervention and control groups at baseline. The final T4 procedures were completed in March 1998.